The conversion of oxygenates to olefins (OTO) is currently the subject of intense research because it has the potential for replacing the long-standing steam cracking technology that is today the industry-standard for producing world scale quantities of ethylene and propylene. The very large volumes involved suggest that substantial economic incentives exist for alternate technologies that can deliver high throughputs of light olefins in a cost efficient manner. Whereas steam cracking relies on non-selective thermal reactions of naphtha range hydrocarbons at very high temperatures, OTO exploits catalytic and micro-architectural properties of acidic molecular sieves under milder temperature conditions to produce high yields of ethylene and propylene from methanol.
Current understanding of the OTO reactions suggests a complex sequence in which three major steps can be identified: (1) an induction period leading to the formation of an active carbon pool (alkyl-aromatics), (2) alkylation-dealkylation reactions of these active intermediates leading to products, and (3) a gradual build-up of condensed ring aromatics. OTO is therefore an inherently transient chemical transformation in which the catalyst is in a continuous state of change. The ability of the catalyst to maintain high olefin yields for prolonged periods of time relies on a delicate balance between the relative rates at which the above processes take place. The formation of coke-like molecules is of singular importance because their accumulation interferes with the desired reaction sequence in a number of ways. In particular, coke renders the carbon pool inactive, lowers the rates of diffusion of reactants and products, increases the potential for undesired secondary reactions and limits catalyst life.
Over the last two decades, many catalytic materials have been identified as being useful for carrying out the OTO reactions. Crystalline molecular sieves are the preferred catalysts today because they simultaneously address the acidity and morphological requirements for the reactions. Some of the most useful molecular sieves for oxygenate converting reactions are the aluminophosphates and metalloaluminophosphates, such as the silicoaluminophosphates (SAPO's). There are a wide variety of SAPO molecular sieves known in the art, of these the more important examples include SAPO-5, SAPO-11, SAPO-18, SAPO-34, SAPO-35, SAPO-41, and SAPO-56. For the OTO process, SAPO molecular sieves having the CHA framework and especially SAPO-34 are particularly important catalysts.
The CHA framework type has a double six-ring structure in an ABC stacking arrangement. The pore openings of the structure are defined by eight member rings that have a diameter of about 4.0 Å, and cylindrical cages within the structure of approximately 10×6.7 Å (“Atlas of Zeolite Framework Types”, 2001, 5 th Edition, p. 96). Other SAPO molecular sieves of the CHA framework type include SAPO-44, SAPO-47 and ZYT-6.
The synthesis of SAPO molecular sieves is a complicated process. There are a number of variables that need to be controlled in order to optimize the synthesis in terms of purity, yield and quality of the SAPO molecular sieve produced. A particularly important variable is the choice of synthesis template, which usually determines which SAPO framework type is obtained from the synthesis. U.S. Pat. No. 4,310,440 teaches that “not all templating agents suitably employed in the preparation of certain species . . . are suitable for the preparation of all members of the generic class.” It is also well known that the same template may induce the formation of different framework types.
In U.S. Pat. No. 4,440,871, the synthesis of a wide variety of SAPO materials of various framework types are described with a number of specific examples. Also disclosed are a large number of possible organic templates, with some specific examples. In the specific examples a number of CHA framework type materials are described. The preparation of SAPO-34 is reported, using tetraethylammonium hydroxide (TEAOH), or isopropylamine, or mixtures of TEAOH and dipropylamine (DPA) as templates. Also disclosed in this patent is a specific example that utilizes cyclohexylamine in the preparation of SAPO44. Although other template materials are described in this patent there are no other templates indicated as being suitable for preparing SAPO's of the CHA framework type. Certain aminoalcohols are mentioned, including triethanolamine, N-methyldiethanolamine, N-methylethanolamine, N,N-dimethylethanolamine and N,N-diethylethanolamine, as possible templates for SAPO molecular sieves. Of these materials, N,N-diethylethanolamine is shown to produce SAPO-5, which is of the AFI framework type. For the other aminoalcohols no indication is provided as to which SAPO or which framework type may be obtained through their use.
Since the synthesis of SAPO-34 was reported in U.S. Pat. No. 4,440,871, tetraethylammonium hydroxide (TEAOH) either alone, or in combination with dipropylamine (DPA), has been the preferred template for preparing SAPO-34. In fact, TEAOH is currently the only quaternary ammonium template known for use in the synthesis of SAPO-34 and other CHA framework type silicoaluminophosphates. However, there are problems associated with the use of TEAOH and DPA. When used alone, TEAOH affords a limited range of synthesis parameters. For example, under certain conditions TEAOH will also template the synthesis of SAPO-18 which has the AEI framework type. TEAOH is thus relatively intolerant to synthesis condition variations. TEAOH is sometimes combined with DPA. However, DPA has a low boiling point (110° C.) resulting in the need for production facilities that can handle high pressures. In certain countries, the use of DPA requires special regulatory authorizations due to its toxicity. Also, DPA is an aggressive template and is often implicated in re-dissolution of the silicoaluminophosphate molecular sieve during its synthesis, resulting in poor quality crystalline product due to surface pitting of the crystals. Finally, it has proved difficult up to now to make pure phase CHA silicoaluminophosphate molecular sieves with a low silica to alumina ratio.
In U.S. Pat. No. 4,440,871, it was reported that SAPO44 was obtained “as the major phase” using cyclohexylamine as a template. In U.S. Pat. No. 6,162,415, relatively pure CHA SAPO-44 was obtained using the same template but with control of the ratio of template to aluminum source and the ratio of phosphorus source to aluminum source. In EP 0 993 867, it was reported that the use of methylbutylamine resulted in SAPO-47 and the use of cyclohexylamine resulted in impure SAPO-44. Methylbutylamine has an even lower boiling point, at 91° C., than DPA.
In the art various attempts have been made to improve the synthesis of AlPO4 and SAPO molecular sieves. One particular desideratum has been lowering the silicon content of the molecular sieve without introducing impurity phases, since it is known that low silicon content has the effect of reducing propane formation and decreasing catalyst deactivation See, for example, Wilson et al., Microporous and Mesoporous Materials, 29, 117-126, 1999. One approach to lowering silicon content has been the addition of a source of fluoride ions to the synthesis mixture.
Thus in U.S. Pat. No. 5,096,684, morpholine and tetraethylammonium hydroxide were found to template the production of SAPO-34 when in the presence of HF. According to this reference, the use of HF in combination with the organic template results in silicoaluminophosphates which have improved thermal and hydrolytic stability. In a Ph.D. thesis (E. H. Halvorsen, University of Oslo, 1996), it was reported that low silica SAPO-34, designated as UiO-S4, was produced using TEAOH template in combination with HF.
However, fluoride based syntheses have the inherent disadvantages that the use of HF presents health and safety problems and requires special processing facilities. There is, therefore, a need for alternative processes for producing pure phase, low silicon, CHA type silicoaluminophosphate molecular sieves and especially for processes which can operate in the absence of fluoride.
U.S. Published Patent Application No. 2003/0232006, published Dec. 18, 2003, discloses a method of synthesizing pure phase CHA framework type silicoaluminophosphate molecular sieves having relatively low silicon content using synthesis templates that contain at least one dimethylamino moiety, selected from one or more of N,N-dimethylethanolamine, N,N-dimethylpropanolamine, N,N-dimethylbutanolamine, N,N-dimethylheptanolamine, N,N-dimethylhexanolamine, N,N-dimethylethylenediamine, N,N-dimethylbutylenediamine, N,N-dimethylheptylenediamine, N,N-dimethylhexylenediamine 1-dimethylamino-2-propanol, N,N-dimethylethylamine, N,N-dimethylpropylamine, N,N-dimethylpentylamine, N,N-dimethylhexylamine and N,N-dimethylheptylamine. The synthesis is conducted in the absence of fluoride.
U.S. Published Patent Application No. 2004/0253163, published Dec. 16, 2004, discloses the synthesis of silicoaluminophosphate molecular sieves having the CHA framework type employing a template having the formula:R1R2N—R3 wherein R1 and R2 are independently selected from the group consisting of alkyl groups having from 1 to 3 carbon atoms and hydroxyalkyl groups having from 1 to 3 carbon atoms and R3 is selected from the group consisting of 4- to 8-membered cycloalkyl groups, optionally substituted by 1 to 3 alkyl groups having from 1 to 3 carbon atoms; and 4- to 8-membered heterocyclic groups having from 1 to 3 heteroatoms, said heterocyclic groups being optionally substituted by 1 to 3 alkyl groups having from 1 to 3 carbon atoms and the heteroatoms in said heterocyclic groups being selected from the group consisting of O, N, and S. Preferably, the template is selected from N,N-dimethyl-cyclohexylamine, N,N-dimethyl-methyl-cyclohexylamine, N,N-dimethyl-cyclopentylamine, N,N-dimethyl-methyl-cyclopentylamine, N,N-dimethyl-cycloheptylamine, N,N-dimethyl-methylcycloheptylamine, and most preferably is N ,N-dimethyl-cyclohexylamine. The synthesis can be effected with or without the presence of fluoride ions and can produce CHA framework type silicoaluminophosphates with a low Si/Al ratio.